A collision detection method for an interventional surgical robot, a robot and a medium

Abstract

The application discloses a collision detection method for an interventional surgery robot, a robot and a medium, and the collision detection method comprises the following steps: a three-dimensional collision detection model of a digital model comprising a robot and a target body is established; a detection point sequence is determined on a pre-planned robot motion track, and a latter detection point in the sequence is farther away from a starting point of the motion track than a former detection point; whether the robot will collide when moving to the position of each detection point in the detection point sequence according to the motion track is determined by using the three-dimensional collision detection model and a bounding box collision detection method; if a collision occurs, a current detection point is replaced by a new detection point based on a preset detection point replacement method, and the detection point sequence is correspondingly updated; and for the updated detection point sequence, collision detection is performed on the replaced detection point and the detection points after the replaced detection point one by one. The collision detection method provided by the application does not need to integrate sensors on the interventional surgery robot, and has low cost and high detection precision and efficiency.

Description

Technical Field

[0001] This invention relates to the field of interventional robotics, and more particularly to a collision detection method, robot, and medium for interventional surgical robots. Background Technology

[0002] Interventional surgical robots are high-precision, high-accuracy, and high-reliability robotic systems designed to support surgeons in performing minimally invasive surgical procedures. Compared to traditional surgical methods, interventional surgical robots offer smaller incisions, more precise procedures, less bleeding, and faster recovery times. In the medical setting, interventional surgical robots have become an increasingly popular technology. Currently, robot collision detection is a hot research area in robotics development and is widely used in the medical field.

[0003] Robot collision detection is a crucial area of ​​modern robotics, ensuring the safety and stability of robots during operation. Robots are indispensable in many modern manufacturing sectors, such as automobile manufacturing, electronics manufacturing, and interventional surgery. During robot operation, collisions can not only damage the robot but also potentially cause serious personal injury and property loss. Therefore, research into robot collision detection technology is increasingly important. Robot collision detection primarily uses sensors to acquire motion and environmental information from the robot, performs feedback control, and issues alarm signals or stops the robot's operation to ensure safety. With the continuous development of artificial intelligence and machine learning technologies, robot collision detection technology is constantly improving and optimizing.

[0004] Traditional robot collision detection methods determine whether a collision has occurred only when the robot moves to or near its current position, and are generally sensor-based. For example, Chinese Patent CN111693040B discloses a robot collision detection method that uses the robot's pose at the output end when in contact with the environment instead of directly measuring the contact torque. When performing dynamic modeling of the robot, it uses the pose of the robot's end effector and the pose of the motor output end to calculate the contact torque between the robot and the environment. It uses a series elastic actuator for collaborative robot improvement and collision detection, adding a pose encoder to the reducer to ensure accurate calculation of the robot's real-time pose state; and combining real-time pose analysis with dynamic modeling eliminates the influence of difficult-to-calculate friction, thus improving the accuracy of collision detection.

[0005] For example, Chinese Patent No. CN115351818A discloses a robot collision detection system, method, robot, robot and chip. The robot collision detection system includes a tactile sensor installed on the robot and a controller connected to the tactile sensor. The controller performs collision detection, including determining the robot's collision state based on the sensing signal of the tactile sensor. The sensing signal is used to characterize the force information in at least one direction to improve the accuracy of collision detection and reduce the probability of collision damage.

[0006] The sensor-based collision detection methods described above determine whether a collision has occurred only when the robot moves to or approaches its current position, resulting in low detection efficiency. Furthermore, the sensor-based robot collision detection methods are highly dependent on the manufacturing and assembly precision of both the robot and the sensors, which not only affects the robot's collision detection accuracy but also makes the sensors very expensive.

[0007] The above background information is provided only to assist in understanding the inventive concept and technical solution of this invention. It does not necessarily belong to the prior art of this application, nor does it necessarily provide technical teaching. In the absence of clear evidence that the above information was disclosed before the filing date of this patent application, the above background information should not be used to evaluate the novelty and inventiveness of this application. Summary of the Invention

[0008] The purpose of this invention is to provide a collision detection method, robot, and medium for interventional surgical robots. The collision detection method for interventional surgical robots provided by this invention does not require the integration of sensors on the interventional surgical robot, resulting in low cost and high detection accuracy and efficiency.

[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0010] A collision detection method for interventional surgical robots includes the following steps:

[0011] Collect spatial environment information for collision detection and establish a three-dimensional collision detection model. The three-dimensional collision detection model includes at least digital models and spatial position information of the target body to be operated on and the robot.

[0012] Based on the pre-planned robot motion trajectory, a sequence of detection points {A1, A2, ..., A...} is determined. n Each detection point in the detection point sequence is located on the motion trajectory, and each subsequent detection point is farther away from the starting point of the motion trajectory than the previous detection point.

[0013] Based on the aforementioned 3D collision detection model, the bounding box collision detection method is used to determine whether a collision will occur when the robot moves to each detection point in the sequence of detection points according to the motion trajectory. If the robot does not collide at the current detection point, collision detection is performed at the next detection point. If the robot does collide at the current detection point, the following steps are performed:

[0014] Based on the preset detection point replacement method, the current detection point A is... i Replace with detection point A′ i The detection point A′ i and the detection point A i The three-dimensional coordinates are different, and the detection point sequence is updated accordingly;

[0015] For the updated detection point sequence, starting from detection point A′ i Initially, for the detection point A′ i The robot will then conduct collision detection at each of the subsequent detection points.

[0016] Furthermore, following any one or a combination of the aforementioned technical solutions, the method further includes, after each update of the detection point sequence, using each detection point in the updated detection point sequence as a control point, and replanning the robot's motion trajectory using a Sel curve motion planning algorithm.

[0017] Furthermore, following any one or a combination of the aforementioned technical solutions, the method further includes, after completing robot collision detection at all detection point positions in the detection point sequence, using each detection point in the latest detection point sequence as control points, and replanning the robot's motion trajectory using a Sel curve motion planning algorithm.

[0018] Furthermore, following any one or a combination of the aforementioned technical solutions, the preset detection point replacement method includes:

[0019] The collision points and their three-dimensional coordinates are determined based on the collision detection results, and the number of collision points is one or more.

[0020] For each of the aforementioned collision points, the following steps are performed:

[0021] Starting from the collision point and taking the current detection point A as the starting point i Determine the first spatial vector for the endpoint;

[0022] Taking the current detection point A as an example i Starting from the first spatial vector, a second spatial vector is determined in the direction of the first spatial vector, and the magnitude of the second spatial vector is a preset first threshold.

[0023] Determine the endpoint of the second spatial vector and its three-dimensional coordinates;

[0024] If there is only one collision point, then the endpoint of the corresponding second spatial vector is determined to be the detection point A′. i ;

[0025] If there are multiple collision points, the target endpoint is determined based on the endpoints of the multiple second spatial vectors, and the target endpoint is taken as the detection point A′. i .

[0026] Furthermore, following any one or a combination of the aforementioned technical solutions, the first threshold is a variable; the more collision points there are, the larger the first threshold becomes, and the fewer collision points there are, the smaller the first threshold becomes.

[0027] Furthermore, following any one or a combination of the aforementioned technical solutions, the preset detection point replacement method includes:

[0028] Based on the current detection point A i Determine the point with the smallest distance from the detection point to the target body to be operated on, and use this point as the starting point, with the detection point A as the starting point. i Determine the third spatial vector for the endpoint;

[0029] With the detection point A i Starting from the third spatial vector, a fourth spatial vector is determined in the direction of the third spatial vector, and the magnitude of the fourth spatial vector is a preset second threshold.

[0030] Determine the endpoint of the fourth spatial vector, and use the endpoint of the fourth spatial vector as the detection point A′. i .

[0031] Furthermore, following any one or a combination of the aforementioned technical solutions, the preset detection point replacement method includes:

[0032] Determine the current detection point A i and its spatial location information;

[0033] The detection point A i The detection point A′ is obtained by translating it vertically upward by a preset distance. i .

[0034] Furthermore, following any one or a combination of the aforementioned technical solutions, based on the three-dimensional collision detection model, the bounding box collision detection method is used to determine whether a collision will occur when the robot moves to each detection point in the detection point sequence according to the motion trajectory, including the following steps:

[0035] Using a pre-established robot kinematic model, the robot's end-effector pose x and displacement velocity are... The angles q of each joint of the robot are calculated.

[0036] Based on the joint angles of the robot, the OBB bounding box of each joint of the robot is established using the OBB bounding box algorithm.

[0037] Based on the aforementioned 3D collision detection model, the OBB bounding box of the target body is established using the OBB bounding box algorithm.

[0038] Based on the OBB bounding boxes of each joint of the robot and the OBB bounding box of the target body, it is determined whether the robot has collided at its current position.

[0039] Furthermore, following any one or a combination of the aforementioned technical solutions, a pre-established robot kinematic model is used to determine the robot's end-effector pose x and displacement velocity. The robot's joint angles q are calculated, including:

[0040] Pre-establish robot kinematics model in, For the robot's joint angular velocity, Let J be the robot joint displacement velocity, and J be the Jacobian matrix;

[0041] Based on the robot's kinematic model, the robot's joint angular velocities are calculated as follows: Among them, J -1 It is the Jacobian inverse matrix;

[0042] Calculate the joint state of the robot at time t as follows: Where q(0) is the initial joint state of the robot, q(t) is the joint state of the robot at time t, and the pose of the robot end effector is the state of each joint of the robot.

[0043] The robot's joint step length Δθ is determined based on the robot's displacement velocity and joint state, and the coordinates and joint angles of each joint of the robot at time t are calculated based on the robot's initial joint angles and the joint step length Δθ.

[0044] Furthermore, following any one or a combination of the aforementioned technical solutions, a three-dimensional rendering and display device is used to render and display the three-dimensional collision detection model, the motion trajectory, and the collision detection results of the robot.

[0045] Furthermore, following any one or a combination of the aforementioned technical solutions, in the detection point sequence {A1, A2, ..., A...} nIn the diagram, A1 is the starting point of the motion trajectory, A n The endpoint of the motion trajectory is the target point for the surgical procedure to be performed.

[0046] Furthermore, following any one or a combination of the aforementioned technical solutions, the motion trajectory of the robot end effector is pre-planned using a Bézier curve motion planning algorithm.

[0047] According to another aspect of the present invention, the present invention provides an interventional robot configured to detect whether a collision will occur when it moves along a pre-planned motion trajectory using the steps of the method described in any one or a combination of the above technical solutions.

[0048] Furthermore, following any one or a combination of the aforementioned technical solutions, the method further includes determining whether an updated detection point sequence exists based on the collision detection results of the robot. If it exists, the robot's motion trajectory is replanned using each detection point in the updated detection point sequence as a control point, and the robot is controlled to move according to the replanned motion trajectory.

[0049] According to another aspect of the present invention, a computer-readable storage medium is provided for storing program instructions configured to be invoked to perform the steps of the method described in any one of the above technical solutions or a combination of multiple technical solutions.

[0050] The beneficial effects of the technical solution provided by this invention are as follows:

[0051] a. This invention establishes a three-dimensional collision detection model including a robot digital model and a target digital model, and determines a sequence of detection points on a pre-planned motion trajectory. It uses bounding box detection to determine whether the robot will collide with other objects during its movement along the pre-planned motion trajectory. This method does not require the integration of multiple sensors into the robot, is low in cost and has good versatility. Moreover, this method does not require the robot to move to the corresponding position to determine the collision detection result, resulting in high collision detection efficiency.

[0052] b. This invention uses a dynamic judgment method for traversal calculation, which decomposes the complex collision problem of the robot into a collision detection sub-problem of each detection point. When a collision occurs in the robot motion simulation, the next traversal only performs collision detection on the updated detection point and subsequent detection points, reducing unnecessary collision calculations and greatly improving the collision detection efficiency of the robot.

[0053] c. After the detection point or detection point sequence is updated, the present invention uses the new detection point to replan the robot's motion trajectory. This can achieve the planning of a new motion trajectory that avoids collisions while detecting whether the robot will collide, assisting the surgeon in planning the best needle insertion route, improving the efficiency of the robot in performing interventional needle insertion, and ensuring the best needle insertion posture.

[0054] d. This invention provides a variety of detection point replacement methods, which can meet the requirements of high collision detection accuracy as well as low computational load and high detection efficiency. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 A flowchart of a collision detection method for an interventional surgical robot, provided as an exemplary embodiment of the present invention;

[0057] Figure 2 A comparison chart showing the detection times required for two different collision detection methods provided as an exemplary embodiment of the present invention;

[0058] Figure 3 A simulation interface diagram of a three-dimensional collision detection model provided for an exemplary embodiment of the present invention. Detailed Implementation

[0059] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0060] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0061] Currently, the end effectors of robots used in interventional surgery are generally small in size. Due to these size limitations, as well as the limitations in the manufacturing and assembly precision of the robot components, integrating multiple sensors into the end effector is very difficult and cannot guarantee optimal detection accuracy. Therefore, existing robot collision detection methods have limitations in their application in interventional robots. Furthermore, the urgency and zero-tolerance requirements of surgical procedures place even higher demands on collision detection in interventional robots. Based on this, this application proposes a robot collision detection method that does not require the integration of sensors into the end effector and offers high detection efficiency and accuracy.

[0062] In one embodiment of the invention, a collision detection method for an interventional surgical robot is provided, see [link to relevant documentation]. Figure 1 This includes the following steps:

[0063] Collect spatial environment information for collision detection and establish a three-dimensional collision detection model. The three-dimensional collision detection model includes digital models and spatial position information of the target body to be operated on, the robot, and other objects. In this embodiment, the other objects refer to other objects such as operating tables and medical instruments that may collide with the robot and are located within the spatial environment information for collision detection. If there are no other objects that may collide with the robot, it is not necessary to establish digital models of other objects.

[0064] Based on the pre-planned robot motion trajectory, a sequence of detection points {A1, A2, ..., A...} is determined. n In the detection point sequence, each detection point is located on the motion trajectory, and each subsequent detection point is farther from the starting point of the motion trajectory than the previous detection point; preferably, A1 is the starting point of the motion trajectory, A n The endpoint of the motion trajectory is the target point for the surgical procedure to be performed.

[0065] Based on the aforementioned 3D collision detection model, the bounding box collision detection method is used to determine whether a collision will occur when the robot moves to each detection point in the sequence of detection points according to the motion trajectory. If the robot does not collide at the current detection point, collision detection is performed at the next detection point. If the robot does collide at the current detection point, the following steps are performed:

[0066] Based on the preset detection point replacement method, the current detection point A is... i Replace with detection point A′ i The detection point A′ i and the detection point A i The three-dimensional coordinates are different, and the detection point sequence is updated accordingly;

[0067] For the updated detection point sequence, starting from detection point A′ i Initially, for the detection point A′ i The robot will then conduct collision detection at each of the subsequent detection points.

[0068] This invention establishes a 3D collision detection model including a digital model of the robot and a digital model of the target object, and determines a sequence of detection points along a pre-planned motion trajectory. It then uses bounding box detection to determine whether the robot will collide with other objects while moving along the pre-planned trajectory. This method does not require integrating multiple sensors onto the robot, resulting in low cost; furthermore, it does not require the robot to reach a specific position to determine the collision detection result, thus achieving high collision detection efficiency.

[0069] Furthermore, traditional collision detection methods traverse bounding boxes layer by layer, finding joints that intersect with other bounding boxes before performing collision detection. These methods increase unnecessary calculations and comparisons, resulting in low detection efficiency. This invention employs a dynamic judgment method for traversal calculations, decomposing the complex robot collision problem into a sub-problem of collision detection at each path point. When a collision occurs during robot motion simulation, the k-th path point in the j-th group is recorded and dynamically updated. Subsequent traversals only detect the k-th path point in the j-th group and subsequent path points that may collide, reducing unnecessary collision calculations. For the same motion trajectory and collision detection environment, five detection points are determined on the motion trajectory to form the detection point sequence. This sequence is then calculated using both traditional layer-by-layer bounding box collision detection and the dynamic judgment method provided in this application. The number of detection points in the detection point sequence is then changed to 10, 15, and 20 points, and compared and verified respectively. The comparison and verification results are as follows: Figure 2As shown, the dynamic judgment method provided in this application reduces the collision detection time by about 85% compared with the traditional method of traversing bounding boxes layer by layer, which greatly improves the collision detection efficiency of the robot.

[0070] In this embodiment, there are multiple methods for establishing the three-dimensional collision detection model. For example, a three-dimensional image of the spatial environment to be collided can be obtained by acquiring one or more of depth images, black-and-white images, and color images of the spatial environment to be collided with using an image acquisition device, and the three-dimensional collision detection model can be established using the three-dimensional image information. Alternatively, a digital model of the robot, a digital model of the target body, and digital models of other objects that may collide with the robot can be directly established, and the positional relationship between each digital model can be determined based on the actual surgical position, thereby forming the three-dimensional collision detection model. Alternatively, a collision detection environment can be pre-set, and a digital human body constructed using CT images and a robot model generated by VTK can be loaded into the collision detection environment, and the three-dimensional collision detection model can be determined based on their relative positions. CT images of the target body, usually the human body to be operated on, are taken using a CT scanner, and a digital human body is generated using the human body CT images. A three-dimensional model of the robot is then established using three-dimensional modeling software. Alternatively, a combination of the above methods can be used to establish the three-dimensional collision detection model. Preferably, a three-dimensional rendering and display device, such as a VTK component, is used to render and display the three-dimensional collision detection model, the motion trajectory, and the collision detection results of the robot.

[0071] In one specific embodiment of the present invention, the simulation interface of the established three-dimensional collision detection model is as follows: Figure 3 As shown. When the VTK component reads the robot's STL type (a 3D software format) model and the CT-generated digital human body, it displays them according to the set coordinate positions. Therefore, the position and size of the robot and the digital human body are determined, and the pre-planned robot motion trajectory is also determined and can be realized. Thus, the entire robot motion process can be simulated and displayed.

[0072] In this embodiment, a pre-established robot kinematic model is used to determine the robot's end-effector pose x and displacement velocity. The joint angles q of the robot are calculated. This represents the pose and displacement velocity of the robot's end effector at the target position. Substituting the values ​​into the kinematic equations, the angles of each joint of the robot are solved by solving a system of equations. This process typically requires mathematical tools, such as matrix operations and iterative methods. It should be noted that in this invention, the robot includes all its joints and the end effector; the objects of study in the robot kinematic model and bounding box detection both include the robot's joints and the end effector.

[0073] Pre-establish robot kinematics model in, For the robot's joint angular velocity, Let J be the displacement velocity of the robot's end effector, and J be the Jacobian matrix.

[0074] right Solving for the given information yields: Among them, J -1 It is the Jacobian inverse matrix.

[0075] The initial joint state of the robot is q(0). The final target joint position can be calculated by integrating the velocity over time. The joint state of the robot at time t is calculated as follows:

[0076] Where q(t) represents the joint state of the robot at time t, and the pose of the robot's end effector represents the state of each joint of the robot.

[0077] This is achieved by discretizing time using the Euler integral method. Given an integration interval Δt, if t k The joint position and velocity at time t are known. k+1 =t k The joint position at time +Δt can be achieved as follows:

[0078]

[0079] Therefore, the kinematic solution focuses on solving for Δθ, i.e.

[0080] Based on the concept of differential kinematics, the motion component Δx of the robot at the target point (TCP) can be approximated by the superposition of the joint motions Δq. The smaller Δx is, the more accurate the linear relationship. That is, assuming the robot starts from the initial point q... s Move to target point q e Then the robot joint step length is obtained as Δθ:

[0081] Δθ=J -1 Δx

[0082] By adding a variable α, which is a gain coefficient, and adjusting the step size to speed up the solution, the iterative kinematic formula is finally obtained:

[0083] Δθ=αJ -1 Δx

[0084] Therefore, the robot's joint step length Δθ is determined based on the robot's displacement velocity and joint state, and the coordinates and joint angles of each joint of the robot at time t are calculated based on the robot's initial joint angles and the joint step length Δθ.

[0085] Based on the aforementioned 3D collision detection model and the aforementioned robot kinematic model, for robot collision detection in each interventional surgery, it is only necessary to import the robot digital model, the human digital model, and the pre-planned motion trajectory into the 3D collision detection model, and use the bounding box detection method to determine whether the robot will collide according to the pre-planned motion trajectory.

[0086] In one embodiment of the present invention, based on the three-dimensional collision detection model, the bounding box collision detection method is used to determine whether a collision will occur when the robot moves to each detection point in the detection point sequence according to the motion trajectory, including the following steps:

[0087] Using a pre-established robot kinematic model, the robot's end-effector pose x and displacement velocity are... The angles q of each joint of the robot are calculated.

[0088] Based on the joint angles of the robot, the OBB bounding box of each joint of the robot is established using the OBB bounding box algorithm.

[0089] Based on the aforementioned 3D collision detection model, the OBB bounding box of the target object and the OBB bounding boxes of other objects are constructed using the OBB bounding box algorithm.

[0090] The system determines whether a collision has occurred at the robot's current position based on the OBB bounding boxes of each joint of the robot, the OBB bounding box of the target object, and the OBB bounding boxes of other objects. Compared to the AABB bounding box collision detection algorithm in bounding box collision detection methods, the OBB bounding box detection method provides higher accuracy and better reliability.

[0091] In this embodiment, a robot is used as an example to illustrate the principle of building an OBB bounding box. OBB bounding boxes are built based on the robot's structure and components, with a corresponding OBB bounding box built for each joint. The idea is to obtain the eigenvectors, i.e., the principal axes of the OBB bounding box, from the vertices of the object's surface using Principal Component Analysis (PCA). The concept of a covariance matrix is ​​introduced to represent the OBB bounding box.

[0092] cov(X i ,X j )=E[(X i -μ i (X) j -μ j )].........(7)

[0093] Among them, X i and X j The coordinates of the bounding box vertices, μ i and μ jMean value of bounding box vertex coordinates, cov(X) i ,X j X represents i and X j The covariance is E, and the expected value is E.

[0094] The covariance matrix M is represented as:

[0095]

[0096] Here, the elements in matrix M represent the variance of the vertex coordinates.

[0097] The eigenvalues ​​and eigenvectors are obtained by solving the covariance matrix M. The direction of the eigenvector corresponding to the largest eigenvalue is the principal axis direction of the OBB bounding box. Then, the coordinates are projected onto the direction vectors to find the maximum values ​​of x and y in each direction. and minimum value Then the center O of the OBB bounding box and the half-length P of the bounding box can be calculated. k :

[0098]

[0099]

[0100] Where k = 0 or k = 1.

[0101] In one embodiment of the present invention, the collision detection method for an interventional surgical robot further includes, after each update of the detection point sequence, using each detection point in the updated detection point sequence as control points, and replanning the robot's motion trajectory using a Sel curve motion planning algorithm. In this embodiment, the robot's motion trajectory is replanned using new detection points after each update of the detection point / detection point sequence, thus enabling the planning of a new motion trajectory that avoids collisions while simultaneously detecting whether the robot will collide.

[0102] In one embodiment of the present invention, the collision detection method for the interventional surgical robot further includes, after completing robot collision detection at all detection point positions in the detection point sequence, using each detection point in the latest detection point sequence as control points, and replanning the robot's motion trajectory using a Sel curve motion planning algorithm. That is, this embodiment does not replan the robot's motion trajectory every time the detection points / detection point sequence is updated, but rather plans a new motion trajectory using the latest detection point sequence after completing robot collision detection at all detection point positions, which can further improve the efficiency of robot collision detection and motion trajectory planning.

[0103] The collision detection method for interventional surgical robots provided by this invention includes a preset detection point replacement method, which is also one of the inventive points. This invention proposes three preset detection point replacement methods, which are described in detail below.

[0104] In one embodiment of the present invention, the preset detection point replacement method includes:

[0105] The collision points and their three-dimensional coordinates are determined based on the collision detection results, and the number of collision points is one or more.

[0106] For each of the aforementioned collision points, the following steps are performed:

[0107] Starting from the collision point and taking the current detection point A as the starting point i Determine the first spatial vector for the endpoint;

[0108] Taking the current detection point A as an example i Starting from the first spatial vector, a second spatial vector is determined in the direction of the first spatial vector. The magnitude of the second spatial vector is a preset first threshold. Preferably, the first threshold is a variable. The more collision points there are, the larger the first threshold is, and the fewer collision points there are, the smaller the first threshold is.

[0109] Determine the endpoint of the second spatial vector and its three-dimensional coordinates;

[0110] If there is only one collision point, then the endpoint of the corresponding second spatial vector is determined to be the detection point A′. i ;

[0111] If there are multiple collision points, the target endpoint is determined based on the endpoints of the multiple second spatial vectors, and the target endpoint is taken as the detection point A′. i Specifically, the target endpoint can be determined by taking the center point of the endpoints of multiple second spatial vectors, or by taking the average coordinates of the endpoints of multiple second spatial vectors, or by taking the endpoints of multiple second spatial vectors that are furthest from the detection point A′. i The farthest point is taken as the target endpoint.

[0112] The detection point A′ determined using the detection point replacement method provided in this embodiment i The detection point A′ must be located far away from the current collision point. Furthermore, the more collision points there are, the larger the current collision range is considered to be. i Distance from the detection point A i The distance is greater, thereby improving the newly determined detection point A′. iThe location-based robot collision detection result is the probability of no collision, and it can also improve the computational efficiency of collision detection.

[0113] In another embodiment of the present invention, the preset detection point replacement method includes:

[0114] Based on the current detection point A i Determine the point with the smallest distance from the detection point to the target body to be operated on, and use this point as the starting point, with the detection point A as the starting point. i Determine the third spatial vector for the endpoint;

[0115] With the detection point A i Starting from the third spatial vector, a fourth spatial vector is determined in the direction of the third spatial vector, and the magnitude of the fourth spatial vector is a preset second threshold.

[0116] Determine the endpoint of the fourth spatial vector, and use the endpoint of the fourth spatial vector as the detection point A′. i .

[0117] The detection point A′ determined by the detection point replacement method provided in this embodiment i It must also be far from the current collision point, and the detection point A′ can be determined relatively quickly. i It should be noted that the target object in this embodiment can also be replaced with other objects that may collide with the robot. If other objects may collide with the robot, the detection point A′ can be determined using the same method. i If the robot collides with more than one object, a replacement detection point can be determined for each object, and the center point of these multiple points can be taken as the detection point A′. i .

[0118] In another embodiment of the present invention, the preset detection point replacement method includes: determining the current detection point A. i and its spatial location information; the detection point A i The detection point A′ is obtained by translating it vertically upward by a preset distance. i In this embodiment, it is not necessary to calculate the collision point or determine the point with the minimum distance from the detection point to the target body to be operated on, nor the first vector to the fourth vector. Therefore, the computational load is very small, and the detection efficiency is very high. Since in interventional surgery, the robot moves towards the human body from far to near and from high to low, this embodiment has great practical significance. For unconventional surgeries where the robot does not move towards the human body from far to near and from high to low, the above two detection point replacement methods can be used.

[0119] In another embodiment of the invention, an interventional robot is provided, the interventional robot being configured to detect whether a collision will occur when it moves along a pre-planned motion trajectory using the steps of the method described in any of the above embodiments.

[0120] In this embodiment, based on the collision detection results of the robot, it is determined whether there is an updated detection point sequence. If there is, each detection point in the updated detection point sequence is used as a control point, and the motion trajectory of the robot is replanned using the Sel curve motion planning algorithm. The robot is then controlled to move according to the replanned motion trajectory.

[0121] In another embodiment of the invention, a computer-readable storage medium is provided for storing program instructions configured to be invoked to perform the steps of the method described in any of the above embodiments.

[0122] It should be noted that the above-described embodiments of the interventional robot and computer-readable storage medium and the embodiments of the collision detection method for interventional surgical robots belong to the same inventive concept. The entire contents of the collision detection method for interventional surgical robots are incorporated into the embodiments of the interventional robot and computer-readable storage medium by reference.

[0123] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0124] The above description is only a specific embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A collision detection method for interventional surgical robots, characterized in that, Includes the following steps: Collect spatial environment information for collision detection and establish a three-dimensional collision detection model. The three-dimensional collision detection model includes at least digital models and spatial position information of the target body to be operated on and the robot. Based on the pre-planned robot motion trajectory, a sequence of detection points {A1, A2, ..., A...} is determined. n Each detection point in the detection point sequence is located on the motion trajectory, and each subsequent detection point is farther away from the starting point of the motion trajectory than the previous detection point. Based on the aforementioned 3D collision detection model, the bounding box collision detection method is used to determine whether a collision will occur when the robot moves to each detection point in the sequence of detection points according to the motion trajectory. If the robot does not collide at the current detection point, collision detection is performed at the next detection point. If the robot does collide at the current detection point, the following steps are performed: Based on the preset detection point replacement method, the current detection point A is... i Replace with detection points The detection point and the detection point A i The three-dimensional coordinates are different, and the detection point sequence is updated accordingly; For the updated detection point sequence, from the detection point Initially, for the detection points The robot will then conduct collision detection at each of the subsequent detection points.

2. The collision detection method for interventional surgical robots according to claim 1, characterized in that, It also includes replanning the robot's motion trajectory using a Sel curve motion planning algorithm, with each detection point in the updated detection point sequence as a control point after each update.

3. The collision detection method for interventional surgical robots according to claim 1, characterized in that, It also includes, after completing robot collision detection at all detection point positions in the detection point sequence, using each detection point in the latest detection point sequence as control points, and replanning the robot's motion trajectory using the Sel curve motion planning algorithm.

4. The collision detection method for interventional surgical robots according to claim 1, characterized in that, The preset detection point replacement method includes: The collision points and their three-dimensional coordinates are determined based on the collision detection results, and the number of collision points is one or more. For each of the aforementioned collision points, the following steps are performed: Starting from the collision point and taking the current detection point A as the starting point i Determine the first spatial vector for the endpoint; Taking the current detection point A as an example i Starting from the first spatial vector, a second spatial vector is determined in the direction of the first spatial vector, and the magnitude of the second spatial vector is a preset first threshold. Determine the endpoint of the second spatial vector and its three-dimensional coordinates; If there is only one collision point, then the endpoint of the corresponding second spatial vector is determined as the detection point. ; If there are multiple collision points, the target endpoint is determined based on the endpoints of the multiple second spatial vectors, and the target endpoint is used as the detection point. .

5. The collision detection method for interventional surgical robots according to claim 4, characterized in that, The first threshold is a variable; the more collision points there are, the larger the first threshold becomes, and the fewer collision points there are, the smaller the first threshold becomes.

6. The collision detection method for an interventional surgical robot according to claim 1, characterized in that, The preset detection point replacement method includes: Based on the current detection point A i Among all detection points, the point with the smallest distance to the target body is determined, and this point is used as the starting point, with detection point A as the reference point. i Determine the third spatial vector for the endpoint; With the detection point A i Starting from the third spatial vector, a fourth spatial vector is determined in the direction of the third spatial vector, and the magnitude of the fourth spatial vector is a preset second threshold. Determine the endpoint of the fourth spatial vector, and use the endpoint of the fourth spatial vector as the detection point. .

7. The collision detection method for an interventional surgical robot according to claim 1, characterized in that, The preset detection point replacement method includes: Determine the current detection point A i and its spatial location information; The detection point A i The detection point is obtained by translating it vertically upward by a preset distance. .

8. The collision detection method for an interventional surgical robot according to claim 1, characterized in that, Based on the aforementioned 3D collision detection model, the bounding box collision detection method is used to determine whether a collision will occur when the robot moves to each detection point in the sequence of detection points according to the motion trajectory. This includes the following steps: Using a pre-established robot kinematics model, the robot's end-effector pose is determined. and displacement velocity The joint angles of the robot were calculated. ; Based on the joint angles of the robot, the OBB bounding box of each joint of the robot is established using the OBB bounding box algorithm. Based on the aforementioned 3D collision detection model, the OBB bounding box of the target body is established using the OBB bounding box algorithm. Based on the OBB bounding boxes of each joint of the robot and the OBB bounding box of the target body, it is determined whether the robot has collided at its current position.

9. The collision detection method for an interventional surgical robot according to claim 8, characterized in that, Using a pre-established robot kinematics model, the robot's end-effector pose is determined. and displacement velocity The joint angles of the robot were calculated. ,include: Pre-establish robot kinematics model ,in, For the robot's joint angular velocity, For the robot's displacement velocity, It is a Jacobian matrix; Based on the robot's kinematic model, the robot's joint angular velocities are calculated as follows: ,in, It is the Jacobian inverse matrix; Calculate the robot in t The joint state at time t is ,in, This represents the initial joint state of the robot. For the robot in t The state of each joint at any given moment; The robot's joint step length is determined based on the robot's displacement velocity and joint state. And based on the robot's initial joint angles and the joint step length calculate t At all times, the robot's various joints and angles.

10. The collision detection method for an interventional surgical robot according to claim 1, characterized in that, The 3D collision detection model, the motion trajectory, and the collision detection results of the robot are rendered and displayed using a 3D rendering and display device.

11. The collision detection method for an interventional surgical robot according to claim 1, characterized in that, In the detection point sequence {A1,A2,……,A} n In the diagram, A1 is the starting point of the motion trajectory, A n The endpoint of the motion trajectory is the target point for the surgical procedure to be performed.

12. The collision detection method for an interventional surgical robot according to claim 1, characterized in that, The motion trajectory of the robot end effector is pre-planned using the Bézier curve motion planning algorithm.

13. An interventional robot, characterized in that, The interventional robot is configured to detect whether a collision will occur when it moves along a pre-planned motion trajectory using the steps of the method as described in any one of claims 1-12.

14. The interventional robot according to claim 13, characterized in that, It also includes determining whether an updated detection point sequence exists based on the collision detection results of the robot. If it exists, the robot's motion trajectory is replanned using each detection point in the updated detection point sequence as a control point, and the robot is controlled to move according to the replanned motion trajectory.

15. A computer-readable storage medium for storing program instructions, characterized in that, The program instructions are configured to be invoked to perform the steps of the method as described in any one of claims 1-12.

Images